Robust and Energy Efficient Temporal Order Event Delivery in Wireless Sensor Networks
Kay Römer · 2002
In wireless sensor networks (WSN), data fusion merges low level distributed sensor events to form high level information, such as speed and direction of a tracked object. As in such examples, data fusion often heavily depends on temporal relationships of sensor events (What happened first? Within which amount of time?). In order for data fusion to produce correct results, it is often very important that data fusion processes events in temporal order (the order in which events happened). Consider for example a temporary partition or delay due to a moving obstacle in the WSN, which keeps part of the events from flowing to the nodes performing data fusion and the effects on the estimated direction and speed of the tracked object. We present to our knowledge the first robust and energy efficient mechanism for temporal order event delivery, which deals with the special requirements of WSN, like energy efficiency and robustness in the face of network dynamics. In simulations this mechanism was up to 20 times more energy efficient than solutions known from classical distributed systems.